December 1, 2013Open Access
Prime Object Proposals with Randomized Prim's Algorithm
Authors
Discussion
Loading...
Member takes
Implication
Randomized trial shows improved object detection rates using a new algorithm in diverse benchmark datasets, suggesting enhanced efficiency.
Key Points
- The aim is to develop an efficient method for generic object detection that outperforms current techniques.
- Introduces a randomized version of Prim's algorithm for object detection.
- Utilizes connectivity graphs of super pixels to generate proposals as bounding-boxes.
- Tests performed on PASCAL VOC 2007, PASCAL VOC 2012, and SUN2012 datasets.
- Achieved 1000 proposals in approximately 0.7 seconds with high detection rates.
- Proposals fit tightly around objects due to their design around super pixel boundaries.
- Outperformed state-of-the-art methods across various evaluation scenarios.
Cite This Study
Manén et al. (2013) studied this question.